TaraAI

Logo

Inspiration

Our team is passionate about advocating for students' rights. One of the main barriers that contributes to college dropout rates is the lack of financial support for transportation, housing, and food. Although public higher education is free in Colombia, poverty remains a major obstacle to successfully completing a degree.

We also realized that there are many educational paths beyond university that are more accessible to people living in poverty and can help them break the cycle of poverty. Then we asked ourselves: what about adults whose main concern is not education, but obtaining the support they need to improve their quality of life?

That is why TaraAI exists.

What it does

TaraAI is an accessible AI tool that helps people identify the government assistance programs they may qualify for based on their personal circumstances. TaraAI asks users about their situation, including their age, city, number of children, and Sisbén score, a Colombian classification system used to prioritize social assistance according to poverty levels.

After gathering this information, TaraAI provides personalized recommendations on the government benefits the user may be eligible for and explains how to access them.

How we built it

We use a n8n agent with conversation chain, free ai models with Groq API, and Buffer memory. We also used Meta API to connect Tara with Whatsapp, inside a Render VPS.

Challenges we ran into

challenges we faced was a lot, we didnt work before on AI agents, deployments and stuff different than personal or university projects, so struggle a lot to put together all of our individual work and make it work, we've learn a lot and even know each other a lot (we werent friends before hackathon, altought study in same university and degree).

Try to get an Meta API and know how to get n8n credentials struggle a lot as well, but we did it.

Accomplishments that we're proud of

We are proud of what we were able to achieve. At the beginning of the project, we had limited knowledge of APIs, AI agents, and collaborative software development. Despite that, we managed to build a functional solution. We had to learn a great deal to turn this idea into reality, but it was definitely worth it.

We are also proud of the impact this project could have. Many people learn about government assistance programs through rumors or incomplete information. TaraAI provides clear and personalized guidance, helping users understand what support they may be eligible for and how to access it.

Another aspect we are proud of is its accessibility. WhatsApp is the most widely used messaging platform in Colombia, making TaraAI easy to use for a large portion of the population. Users do not need to download any additional applications; they simply start a conversation with the chatbot, just as they would with any other contact.

What we learned

weve learned how to use APIs, deploy a repository, chatbot fundamentals, Using docker to run projects quickly, test using AI such as Claude Code, and a rught use of logs.

What's next for TaraAI

We plan to improve agent context to implement RAG system, and then publish it as open source, we are open to receive changes, let people exploit this idea and use this project as a template for new Whatsapp agents projects.